pytrilogy 0.0.1.105__py3-none-any.whl → 0.0.1.107__py3-none-any.whl

{pytrilogy-0.0.1.105.dist-info → pytrilogy-0.0.1.107.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: pytrilogy
-Version: 0.0.1.105
+Version: 0.0.1.107
 Summary: Declarative, typed query language that compiles to SQL.
 Home-page: 
 Author: 
@@ -275,3 +275,81 @@ but all are worth checking out. Please open PRs/comment for anything missed!
 - [malloy](https://github.com/malloydata/malloy)
 - [preql](https://github.com/erezsh/Preql)
 - [PREQL](https://github.com/PRQL/prql)
+
+## Minimal Syntax Reference
+
+#### IMPORT
+
+`import <path> as <alias>;`
+
+#### CONCEPT
+
+Types: `string | int | float | bool | date | datetime | time | timestamp | interval`;
+
+Key:
+`key <name> <type>;`
+
+Property:
+`property <key>.<name> <type>;`
+
+Transformation:
+`auto <name> <- <expression>;`
+
+#### DATASOURCE
+```sql
+datasource <name>(
+    <column>:<concept>,
+    <column>:<concept>,
+)
+grain(<concept>, <concept>)
+address <table>;
+```
+
+#### SELECT
+
+Primary acces
+
+```sql
+select
+    <concept>,
+    <concept>+1 -> <alias>
+WHERE
+    <concept> = <value>
+ORDER BY
+    <concept> asc|desc
+;
+```
+
+#### CTE/ROWSET
+
+Reusable virtual set of rows. Useful for windows, filtering. 
+
+```sql
+with <alias> as
+select
+    <concept>,
+    <concept>+1 -> <alias>
+WHERE
+    <concept> = <value>
+
+select <alias>.<concept>;
+
+```
+
+
+#### PERSIST
+
+Store output of a query in a warehouse table
+
+```sql
+persist <alias> as <table_name> from
+<select>;
+```
+
+#### SHOW
+
+Return generated SQL without executing.
+
+```sql
+show <select>;
+```

{pytrilogy-0.0.1.105.dist-info → pytrilogy-0.0.1.107.dist-info}/RECORD

@@ -1,8 +1,8 @@
-trilogy/__init__.py,sha256=7VUwx55iCGgc689L41AJd9zl_nZZvy9zumsQb-P67xc,245
+trilogy/__init__.py,sha256=ouq-RNu0DVYw8n1C2ekRmcAJ_SL_PFzbRBzo2O814TM,292
 trilogy/compiler.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-trilogy/constants.py,sha256=MphpyZXP4URq7R1vP8EUtT_ZfGPSKuqKKFGMMjfMRtA,482
+trilogy/constants.py,sha256=LxiK2TiVQPEa6tXkxWk9DJHOR3zsGNSqgQuqtOf66cw,518
 trilogy/engine.py,sha256=R5ubIxYyrxRExz07aZCUfrTsoXCHQ8DKFTDsobXdWdA,1102
-trilogy/executor.py,sha256=pQy8Xo7rxtZ9VZoIkOUpOHRzy6eQ5ayZtnBOtf3Zjyg,8359
+trilogy/executor.py,sha256=xF6wzbhP6a3wz4nrxsRCKeKF7qytUQEL75oI3BGJ2hQ,8744
 trilogy/parser.py,sha256=UtuqSiGiCjpMAYgo1bvNq-b7NSzCA5hzbUW31RXaMII,281
 trilogy/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 trilogy/utility.py,sha256=zM__8r29EsyDW7K9VOHz8yvZC2bXFzh7xKy3cL7GKsk,707
@@ -16,8 +16,9 @@ trilogy/core/exceptions.py,sha256=NvV_4qLOgKXbpotgRf7c8BANDEvHxlqRPaA53IThQ2o,56
 trilogy/core/functions.py,sha256=zkRReytiotOBAW-a3Ri5eoejZDYTt2-7Op80ZxZxUmw,9129
 trilogy/core/graph_models.py,sha256=oJUMSpmYhqXlavckHLpR07GJxuQ8dZ1VbB1fB0KaS8c,2036
 trilogy/core/internal.py,sha256=jNGFHKENnbMiMCtAgsnLZYVSENDK4b5ALecXFZpTDzQ,1075
-trilogy/core/models.py,sha256=S2LTFh7F92PUMelY1taXn2BuimSZpYZa7QulXVoUkNM,105556
-trilogy/core/query_processor.py,sha256=x3fjs1Vhg_G1FHhjLomZ3kH16PU9RPcediRz-Lu_QZg,11625
+trilogy/core/models.py,sha256=WH7GHgn1a3xegVo12_NTP7V_ptN-_ObY7s4ZgkPI0D4,108548
+trilogy/core/optimization.py,sha256=SpWRQL1biAUvMCijk2I-FCQY2KzXd4eiu3ZlxzVE-uQ,4505
+trilogy/core/query_processor.py,sha256=w_CS2TEmSk8Bhk6ukxLavQgQyA9UwcgoPChouLREujQ,11747
 trilogy/core/processing/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 trilogy/core/processing/concept_strategies_v3.py,sha256=6Z1ODKbvxaWFMRHa3vX87oSBW9XKtC7S7pUxburCVkA,22369
 trilogy/core/processing/graph_utils.py,sha256=ulCJ4hYAISbUxLD6VM2fah9RBPGIXSEHEPeRBSFl0Rs,1197
@@ -36,24 +37,24 @@ trilogy/core/processing/node_generators/select_node.py,sha256=xeCqIUEubrf3u_QQfb
 trilogy/core/processing/node_generators/unnest_node.py,sha256=s1VXQZSf1LnX3ISeQ5JzmzmCKUw30-5OK_f0YTB9_48,1031
 trilogy/core/processing/node_generators/window_node.py,sha256=ekazi5eXxnShpcp-qukXNG4DHFdULoXrX-YWUWLNEpM,2527
 trilogy/core/processing/nodes/__init__.py,sha256=ZkDGQksvsM5uNia5rhXFCUJcpTRhoYYFdyfJw-Eiu8s,3674
-trilogy/core/processing/nodes/base_node.py,sha256=ovFM4r8QG2sy5d2X1MTCk5BiYUwvr0chxSaIopkO6Fc,8890
+trilogy/core/processing/nodes/base_node.py,sha256=VaK4rWV8PQMyTPTgStmCbPbV0mmTHyTwBhz0C0N2KG0,8961
 trilogy/core/processing/nodes/filter_node.py,sha256=DqSRv8voEajPZqzeeiIsxuv4ubvsmeQcCW6x_v2CmOk,1359
-trilogy/core/processing/nodes/group_node.py,sha256=-xaZuAkCXHWP_K5BmoJ5jTM_53wYWtYPEkrr4LaWib0,3735
-trilogy/core/processing/nodes/merge_node.py,sha256=sXaxKEFdPrTYJyDwcs8lNBoAvz0Nebfd8xcdUtavOoU,12332
+trilogy/core/processing/nodes/group_node.py,sha256=Y_NWB_AwFrE-YithjZ7lYYDN4e0el4su3ICq2EIr3HA,3837
+trilogy/core/processing/nodes/merge_node.py,sha256=pAtRTqkpsfE_pmqPSdeV0rHwFPzclJ3WItRLX8AuuJw,12609
 trilogy/core/processing/nodes/select_node_v2.py,sha256=tAADeVruch-flFiedbY1zi7ukMG2RpWecvxxZ5aL3ZU,6354
 trilogy/core/processing/nodes/unnest_node.py,sha256=t4kY3a_dR3iXistPemStfdw0uJfnxwTcoQg1HiDa3xo,1501
 trilogy/core/processing/nodes/window_node.py,sha256=QjAWgqBZqFSRCPwc7JBmgQJobWW50rsHI0pjJe0Zzg0,926
 trilogy/dialect/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-trilogy/dialect/base.py,sha256=xWSbU50tiG55aDPXGxpUbgg4clGuvtnBBEHFWn7cku8,28669
-trilogy/dialect/bigquery.py,sha256=oiyZ9GejmRJocMFcprMM1ZxTSYI5Po5PXHJKvQSltWs,2845
-trilogy/dialect/common.py,sha256=TI0UZk8N-M-cQbGPw62uUTaxDTnnsH-zJqhPOkD0KtQ,1913
-trilogy/dialect/config.py,sha256=XjxvUmnF_QDc2IE70TnySMsiBuKs3iebC_O0ZJ34atE,1491
-trilogy/dialect/duckdb.py,sha256=vEf6srsjByGS4Xt6ou5jrOj98HOoHaOtZhlsoEo1EIU,3016
-trilogy/dialect/enums.py,sha256=5fDOyQxgBQoUOgzc_ARdtTXNiEhhYDfrk7gc5scMJiw,3402
-trilogy/dialect/postgres.py,sha256=aY9E9gCzprVn1skCI-qdBG6E9HsuRX77XpFk1XbgSTY,3144
-trilogy/dialect/presto.py,sha256=9rUqdx1fCekTSnzhHVSW71LFBW2uWjKSobohkcBA7vw,2728
-trilogy/dialect/snowflake.py,sha256=r_biJMiJdzBwOXp092TKYpaGd8OxbQiHetDGYCGoySI,2879
-trilogy/dialect/sql_server.py,sha256=ood_PftLs3CehH7UGH72WRZ9Ezw6kjdiTtXv4KFInvg,2837
+trilogy/dialect/base.py,sha256=nYrm7Z-GnVVhr5vWHjMghWkGoq2r7ogzoGVasAGTxGo,29223
+trilogy/dialect/bigquery.py,sha256=9vxQn2BMv_oTGQSWQpoN5ho_OgqMWaHH9e-5vQVf44c,2906
+trilogy/dialect/common.py,sha256=zWrYmvevlXznocw9uGHmY5Ws1rp_kICm9zA_ulTe4eg,2165
+trilogy/dialect/config.py,sha256=JdGIiHf2EVoFNTYzqQUy1bMmzqZiFTjcnYglzAMa4dM,3351
+trilogy/dialect/duckdb.py,sha256=Ddyt68sr8IL2HnZMenyytoD65FXwY_O2pz1McyS0bis,3075
+trilogy/dialect/enums.py,sha256=4NdpsydBpDn6jnh0JzFz5VvQEtnShErWtWHVyT6TNpw,3948
+trilogy/dialect/postgres.py,sha256=r47xbCA7nfEYENofiVfLZ-SnReNfDmUmW4OSHVkkP4E,3206
+trilogy/dialect/presto.py,sha256=rOr-ftb0S5AeEncQgp6jj1iCWlCEkzZuQC3TGMUcemg,2790
+trilogy/dialect/snowflake.py,sha256=N3HknYgN-fjD7BLX1Ucj-ss_ku2Ox8DgLsF3BIHutHo,2941
+trilogy/dialect/sql_server.py,sha256=UrLeA9bxiFJ4qpGsqVJqBybQCyJhetMebe8IzQW1q9s,2900
 trilogy/docs/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 trilogy/hooks/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 trilogy/hooks/base_hook.py,sha256=Xkb-A2qCHozYjum0A36zOy5PwTVwrP3NLDF0U2GpgHo,1100
@@ -65,13 +66,13 @@ trilogy/parsing/common.py,sha256=lz0IyVA8v-u-DGFgzkmdb4_00I--Kegmo9HNF7CrajI,579
 trilogy/parsing/config.py,sha256=Z-DaefdKhPDmSXLgg5V4pebhSB0h590vI0_VtHnlukI,111
 trilogy/parsing/exceptions.py,sha256=92E5i2frv5hj9wxObJZsZqj5T6bglvPzvdvco_vW1Zk,38
 trilogy/parsing/helpers.py,sha256=AbpHGcgLb-kRsJGnwFEktk7uzpZOCcBY74-YBdrKVGs,1
-trilogy/parsing/parse_engine.py,sha256=2IK7nAz_X9WiAqWG9cMPsj3IW6bGS4vqSWB4afvyUF4,63845
+trilogy/parsing/parse_engine.py,sha256=TLy56pDatDfzfwbJkrJ-XXB05s_VW9_iRrkwtKR0GR4,63860
 trilogy/parsing/render.py,sha256=fxjpq2FZLgllw_d4cru-t_IXNPAz2DmYkT7v9ED0XRI,11540
 trilogy/scripts/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 trilogy/scripts/trilogy.py,sha256=PHxvv6f2ODv0esyyhWxlARgra8dVhqQhYl0lTrSyVNo,3729
-pytrilogy-0.0.1.105.dist-info/LICENSE.md,sha256=5ZRvtTyCCFwz1THxDTjAu3Lidds9WjPvvzgVwPSYNDo,1042
-pytrilogy-0.0.1.105.dist-info/METADATA,sha256=VTUiVso4QpSImu15q4jsbq0FQ7HY3hUi9GOPan69sfI,6883
-pytrilogy-0.0.1.105.dist-info/WHEEL,sha256=y4mX-SOX4fYIkonsAGA5N0Oy-8_gI4FXw5HNI1xqvWg,91
-pytrilogy-0.0.1.105.dist-info/entry_points.txt,sha256=0petKryjvvtEfTlbZC1AuMFumH_WQ9v8A19LvoS6G6c,54
-pytrilogy-0.0.1.105.dist-info/top_level.txt,sha256=cAy__NW_eMAa_yT9UnUNlZLFfxcg6eimUAZ184cdNiE,8
-pytrilogy-0.0.1.105.dist-info/RECORD,,
+pytrilogy-0.0.1.107.dist-info/LICENSE.md,sha256=5ZRvtTyCCFwz1THxDTjAu3Lidds9WjPvvzgVwPSYNDo,1042
+pytrilogy-0.0.1.107.dist-info/METADATA,sha256=KF68qL9kNj855oUEtUVCCMd5hei-LRfABfoaEWlTa8g,7882
+pytrilogy-0.0.1.107.dist-info/WHEEL,sha256=Z4pYXqR_rTB7OWNDYFOm1qRk0RX6GFP2o8LgvP453Hk,91
+pytrilogy-0.0.1.107.dist-info/entry_points.txt,sha256=0petKryjvvtEfTlbZC1AuMFumH_WQ9v8A19LvoS6G6c,54
+pytrilogy-0.0.1.107.dist-info/top_level.txt,sha256=cAy__NW_eMAa_yT9UnUNlZLFfxcg6eimUAZ184cdNiE,8
+pytrilogy-0.0.1.107.dist-info/RECORD,,

{pytrilogy-0.0.1.105.dist-info → pytrilogy-0.0.1.107.dist-info}/WHEEL

@@ -1,5 +1,5 @@
 Wheel-Version: 1.0
-Generator: setuptools (70.2.0)
+Generator: setuptools (70.3.0)
 Root-Is-Purelib: true
 Tag: py3-none-any
 

trilogy/__init__.py

@@ -2,7 +2,8 @@ from trilogy.core.models import Environment
 from trilogy.dialect.enums import Dialects
 from trilogy.executor import Executor
 from trilogy.parser import parse
+from trilogy.constants import CONFIG
 
-__version__ = "0.0.1.105"
+__version__ = "0.0.1.107"
 
-__all__ = ["parse", "Executor", "Dialects", "Environment"]
+__all__ = ["parse", "Executor", "Dialects", "Environment", "CONFIG"]

trilogy/constants.py

@@ -23,6 +23,7 @@ NULL_VALUE = MagicConstants.NULL
 class Config:
     strict_mode: bool = True
     human_identifiers: bool = True
+    inline_datasources: bool = True
 
 
 CONFIG = Config()

trilogy/core/models.py

@@ -34,7 +34,11 @@ from pydantic import (
 from lark.tree import Meta
 from pathlib import Path
 from trilogy.constants import logger, DEFAULT_NAMESPACE, ENV_CACHE_NAME, MagicConstants
-from trilogy.core.constants import ALL_ROWS_CONCEPT, INTERNAL_NAMESPACE
+from trilogy.core.constants import (
+    ALL_ROWS_CONCEPT,
+    INTERNAL_NAMESPACE,
+    CONSTANT_DATASET,
+)
 from trilogy.core.enums import (
     InfiniteFunctionArgs,
     Purpose,
@@ -1434,6 +1438,7 @@ class MultiSelectStatement(Namespaced, BaseModel):
 
 class Address(BaseModel):
     location: str
+    is_query: bool = False
 
 
 class Query(BaseModel):
@@ -1529,10 +1534,27 @@ class Datasource(Namespaced, BaseModel):
         default_factory=lambda: DatasourceMetadata(freshness_concept=None)
     )
 
+    @property
+    def condition(self):
+        return None
+
     @cached_property
     def output_lcl(self) -> LooseConceptList:
         return LooseConceptList(concepts=self.output_concepts)
 
+    @property
+    def can_be_inlined(self) -> bool:
+        if isinstance(self.address, Address) and self.address.is_query:
+            return False
+        for x in self.columns:
+            if not isinstance(x.alias, str):
+                return False
+        return True
+
+    @property
+    def non_partial_concept_addresses(self) -> set[str]:
+        return set([c.address for c in self.full_concepts])
+
     @field_validator("namespace", mode="plain")
     @classmethod
     def namespace_validation(cls, v):
@@ -1647,7 +1669,7 @@ class Datasource(Namespaced, BaseModel):
         namespace = self.namespace.replace(".", "_") if self.namespace else ""
         return f"{namespace}_{self.identifier}"
 
-    @cached_property
+    @property
     def safe_location(self) -> str:
         if isinstance(self.address, Address):
             return self.address.location
@@ -1746,7 +1768,7 @@ class QueryDatasource(BaseModel):
     input_concepts: List[Concept]
     output_concepts: List[Concept]
     source_map: Dict[str, Set[Union[Datasource, "QueryDatasource", "UnnestJoin"]]]
-    datasources: Sequence[Union[Datasource, "QueryDatasource"]]
+    datasources: List[Union[Datasource, "QueryDatasource"]]
     grain: Grain
     joins: List[BaseJoin | UnnestJoin]
     limit: Optional[int] = None
@@ -1797,10 +1819,7 @@ class QueryDatasource(BaseModel):
             c.address for c in values["input_concepts"]
         )
         seen = set()
-        for k, val in v.items():
-            # if val:
-            #     if len(val) != 1:
-            #         raise SyntaxError(f"source map {k} has multiple values {len(val)}")
+        for k, _ in v.items():
             seen.add(k)
         for x in expected:
             if x not in seen:
@@ -1922,18 +1941,18 @@ class QueryDatasource(BaseModel):
         )
 
     def get_alias(
-        self, concept: Concept, use_raw_name: bool = False, force_alias: bool = False
+        self,
+        concept: Concept,
+        use_raw_name: bool = False,
+        force_alias: bool = False,
+        source: str | None = None,
     ):
-        # if we should use the raw datasource name to access
-        use_raw_name = (
-            True
-            if (len(self.datasources) == 1 or use_raw_name) and not force_alias
-            # if ((len(self.datasources) == 1 and isinstance(self.datasources[0], Datasource)) or use_raw_name) and not force_alias
-            else False
-        )
         for x in self.datasources:
             # query datasources should be referenced by their alias, always
             force_alias = isinstance(x, QueryDatasource)
+            use_raw_name = isinstance(x, Datasource) and not force_alias
+            if source and x.identifier != source:
+                continue
             try:
                 return x.get_alias(
                     concept.with_grain(self.grain),
@@ -1967,8 +1986,7 @@ class Comment(BaseModel):
 
 class CTE(BaseModel):
     name: str
-    source: "QueryDatasource"  # TODO: make recursive
-    # output columns are what are selected/grouped by
+    source: "QueryDatasource"
     output_columns: List[Concept]
     source_map: Dict[str, str | list[str]]
     grain: Grain
@@ -1979,6 +1997,10 @@ class CTE(BaseModel):
     condition: Optional[Union["Conditional", "Comparison", "Parenthetical"]] = None
     partial_concepts: List[Concept] = Field(default_factory=list)
     join_derived_concepts: List[Concept] = Field(default_factory=list)
+    order_by: Optional[OrderBy] = None
+    limit: Optional[int] = None
+    requires_nesting: bool = True
+    base_name_override: Optional[str] = None
 
     @computed_field  # type: ignore
     @property
@@ -1989,6 +2011,40 @@ class CTE(BaseModel):
     def validate_output_columns(cls, v):
         return unique(v, "address")
 
+    def inline_parent_datasource(self, parent: CTE) -> bool:
+        qds_being_inlined = parent.source
+        ds_being_inlined = qds_being_inlined.datasources[0]
+        if not isinstance(ds_being_inlined, Datasource):
+            return False
+        self.source.datasources = [
+            ds_being_inlined,
+            *[
+                x
+                for x in self.source.datasources
+                if x.identifier != qds_being_inlined.identifier
+            ],
+        ]
+        # need to identify this before updating joins
+        if self.base_name == parent.name:
+            self.base_name_override = ds_being_inlined.safe_location
+
+        for join in self.joins:
+            if isinstance(join, InstantiatedUnnestJoin):
+                continue
+            if join.left_cte.name == parent.name:
+                join.left_cte = ds_being_inlined
+            if join.right_cte.name == parent.name:
+                join.right_cte = ds_being_inlined
+        for k, v in self.source_map.items():
+            if isinstance(v, list):
+                self.source_map[k] = [
+                    ds_being_inlined.name if x == parent.name else x for x in v
+                ]
+            elif v == parent.name:
+                self.source_map[k] = ds_being_inlined.name
+        self.parent_ctes = [x for x in self.parent_ctes if x.name != parent.name]
+        return True
+
     def __add__(self, other: "CTE"):
         logger.info('Merging two copies of CTE "%s"', self.name)
         if not self.grain == other.grain:
@@ -2031,16 +2087,25 @@ class CTE(BaseModel):
     def relevant_base_ctes(self):
         return self.parent_ctes
 
+    @property
+    def is_root_datasource(self) -> bool:
+        return (
+            len(self.source.datasources) == 1
+            and isinstance(self.source.datasources[0], Datasource)
+            and not self.source.datasources[0].name == CONSTANT_DATASET
+        )
+
     @property
     def base_name(self) -> str:
+        if self.base_name_override:
+            return self.base_name_override
         # if this cte selects from a single datasource, select right from it
         valid_joins: List[Join] = [
             join for join in self.joins if isinstance(join, Join)
         ]
-        if len(self.source.datasources) == 1 and isinstance(
-            self.source.datasources[0], Datasource
-        ):
+        if self.is_root_datasource:
             return self.source.datasources[0].safe_location
+
         # if we have multiple joined CTEs, pick the base
         # as the root
         elif len(self.source.datasources) == 1 and len(self.parent_ctes) == 1:
@@ -2066,11 +2131,10 @@ class CTE(BaseModel):
 
     @property
     def base_alias(self) -> str:
+
+        if self.is_root_datasource:
+            return self.source.datasources[0].identifier
         relevant_joins = [j for j in self.joins if isinstance(j, Join)]
-        if len(self.source.datasources) == 1 and isinstance(
-            self.source.datasources[0], Datasource
-        ):
-            return self.source.datasources[0].full_name.replace(".", "_")
         if relevant_joins:
             return relevant_joins[0].left_cte.name
         elif self.relevant_base_ctes:
@@ -2079,12 +2143,16 @@ class CTE(BaseModel):
             return self.parent_ctes[0].name
         return self.name
 
-    def get_alias(self, concept: Concept) -> str:
+    def get_alias(self, concept: Concept, source: str | None = None) -> str:
         for cte in self.parent_ctes:
             if concept.address in [x.address for x in cte.output_columns]:
+                if source and source != cte.name:
+                    continue
                 return concept.safe_address
         try:
-            source = self.source.get_alias(concept)
+            source = self.source.get_alias(concept, source=source)
+            if not source:
+                raise ValueError("No source found")
             return source
         except ValueError as e:
             return f"INVALID_ALIAS: {str(e)}"
@@ -2097,6 +2165,11 @@ class CTE(BaseModel):
             and not self.group_to_grain
         ):
             return False
+        if (
+            len(self.source.datasources) == 1
+            and self.source.datasources[0].name == CONSTANT_DATASET
+        ):
+            return False
         return True
 
     @property
@@ -2130,19 +2203,43 @@ class JoinKey(BaseModel):
 
 
 class Join(BaseModel):
-    left_cte: CTE
-    right_cte: CTE
+    left_cte: CTE | Datasource
+    right_cte: CTE | Datasource
     jointype: JoinType
     joinkeys: List[JoinKey]
 
+    @property
+    def left_name(self) -> str:
+        if isinstance(self.left_cte, Datasource):
+            return self.left_cte.identifier
+        return self.left_cte.name
+
+    @property
+    def right_name(self) -> str:
+        if isinstance(self.right_cte, Datasource):
+            return self.right_cte.identifier
+        return self.right_cte.name
+
+    @property
+    def left_ref(self) -> str:
+        if isinstance(self.left_cte, Datasource):
+            return f"{self.left_cte.safe_location} as {self.left_cte.identifier}"
+        return self.left_cte.name
+
+    @property
+    def right_ref(self) -> str:
+        if isinstance(self.right_cte, Datasource):
+            return f"{self.right_cte.safe_location} as {self.right_cte.identifier}"
+        return self.right_cte.name
+
     @property
     def unique_id(self) -> str:
-        return self.left_cte.name + self.right_cte.name + self.jointype.value
+        return self.left_name + self.right_name + self.jointype.value
 
     def __str__(self):
         return (
-            f"{self.jointype.value} JOIN {self.left_cte.name} and"
-            f" {self.right_cte.name} on {','.join([str(k) for k in self.joinkeys])}"
+            f"{self.jointype.value} JOIN {self.left_name} and"
+            f" {self.right_name} on {','.join([str(k) for k in self.joinkeys])}"
         )
 
 

trilogy/core/optimization.py

@@ -0,0 +1,141 @@
+from trilogy.core.models import (
+    CTE,
+    SelectStatement,
+    PersistStatement,
+    Datasource,
+    MultiSelectStatement,
+)
+from trilogy.core.enums import PurposeLineage
+from trilogy.constants import logger
+from abc import ABC
+
+
+class OptimizationRule(ABC):
+
+    def optimize(self, cte: CTE) -> bool:
+        raise NotImplementedError
+
+    def log(self, message: str):
+        logger.info(f"[Optimization][{self.__class__.__name__}] {message}")
+
+
+class InlineDatasource(OptimizationRule):
+
+    def optimize(self, cte: CTE) -> bool:
+        if not cte.parent_ctes:
+            return False
+
+        optimized = False
+        self.log(
+            f"Checking {cte.name} for consolidating inline tables with {len(cte.parent_ctes)} parents"
+        )
+        to_inline: list[CTE] = []
+        for parent_cte in cte.parent_ctes:
+            if not parent_cte.is_root_datasource:
+                self.log(f"parent {parent_cte.name} is not root")
+                continue
+            if parent_cte.parent_ctes:
+                self.log(f"parent {parent_cte.name} has parents")
+                continue
+            raw_root = parent_cte.source.datasources[0]
+            if not isinstance(raw_root, Datasource):
+                self.log(f"parent {parent_cte.name} is not datasource")
+                continue
+            root: Datasource = raw_root
+            if not root.can_be_inlined:
+                self.log(f"parent {parent_cte.name} datasource is not inlineable")
+                continue
+            root_outputs = {x.address for x in root.output_concepts}
+            cte_outputs = {x.address for x in parent_cte.output_columns}
+            if not cte_outputs.issubset(root_outputs):
+                self.log(f"Not all {parent_cte.name} outputs are found on datasource")
+                continue
+
+            to_inline.append(parent_cte)
+
+        for replaceable in to_inline:
+            self.log(f"Inlining parent {replaceable.name}")
+            cte.inline_parent_datasource(replaceable)
+
+        return optimized
+
+
+REGISTERED_RULES: list[OptimizationRule] = [InlineDatasource()]
+
+
+def filter_irrelevant_ctes(input: list[CTE], root_cte: CTE):
+    relevant_ctes = set()
+
+    def recurse(cte: CTE):
+        relevant_ctes.add(cte.name)
+        for cte in cte.parent_ctes:
+            recurse(cte)
+
+    recurse(root_cte)
+    return [cte for cte in input if cte.name in relevant_ctes]
+
+
+def is_direct_return_eligible(
+    cte: CTE, select: SelectStatement | PersistStatement | MultiSelectStatement
+) -> bool:
+    if isinstance(select, (PersistStatement, MultiSelectStatement)):
+        return False
+    derived_concepts = [
+        c for c in cte.source.output_concepts if c not in cte.source.input_concepts
+    ]
+    eligible = True
+    conditions = (
+        set(x.address for x in select.where_clause.concept_arguments)
+        if select.where_clause
+        else set()
+    )
+    if conditions and select.limit:
+        return False
+    for x in derived_concepts:
+        if x.derivation == PurposeLineage.WINDOW:
+            return False
+        if x.derivation == PurposeLineage.AGGREGATE:
+            if x.address in conditions:
+                return False
+    logger.info(
+        f"Upleveling output select to final CTE with derived_concepts {[x.address for x in derived_concepts]}"
+    )
+    return eligible
+
+
+def sort_select_output(cte: CTE, query: SelectStatement | MultiSelectStatement):
+    hidden_addresses = [c.address for c in query.hidden_components]
+    output_addresses = [
+        c.address for c in query.output_components if c.address not in hidden_addresses
+    ]
+
+    mapping = {x.address: x for x in cte.output_columns}
+
+    new_output = []
+    for x in output_addresses:
+        new_output.append(mapping[x])
+    cte.output_columns = new_output
+
+
+def optimize_ctes(
+    input: list[CTE], root_cte: CTE, select: SelectStatement | MultiSelectStatement
+):
+    complete = False
+
+    while not complete:
+        actions_taken = False
+        for rule in REGISTERED_RULES:
+            for cte in input:
+                actions_taken = rule.optimize(cte)
+        complete = not actions_taken
+
+    if is_direct_return_eligible(root_cte, select):
+        root_cte.order_by = select.order_by
+        root_cte.limit = select.limit
+        root_cte.condition = (
+            select.where_clause.conditional if select.where_clause else None
+        )
+        root_cte.requires_nesting = False
+        sort_select_output(cte, select)
+
+    return filter_irrelevant_ctes(input, root_cte)

trilogy/core/processing/nodes/base_node.py

@@ -45,7 +45,7 @@ def concept_list_to_grain(
 
 
 def resolve_concept_map(
-    inputs: List[QueryDatasource],
+    inputs: List[QueryDatasource | Datasource],
     targets: List[Concept],
     inherited_inputs: List[Concept],
     full_joins: List[Concept] | None = None,
@@ -156,7 +156,9 @@ class StrategyNode:
         return f"{self.__class__.__name__}<{contents}>"
 
     def _resolve(self) -> QueryDatasource:
-        parent_sources = [p.resolve() for p in self.parents]
+        parent_sources: List[QueryDatasource | Datasource] = [
+            p.resolve() for p in self.parents
+        ]
 
         # if conditional:
         #     for condition in conditions[1:]:

trilogy/core/processing/nodes/group_node.py

@@ -4,6 +4,7 @@ from trilogy.constants import logger
 from trilogy.core.models import (
     Grain,
     QueryDatasource,
+    Datasource,
     SourceType,
     Concept,
     Environment,
@@ -45,7 +46,9 @@ class GroupNode(StrategyNode):
         )
 
     def _resolve(self) -> QueryDatasource:
-        parent_sources: list[QueryDatasource] = [p.resolve() for p in self.parents]
+        parent_sources: List[QueryDatasource | Datasource] = [
+            p.resolve() for p in self.parents
+        ]
 
         grain = concept_list_to_grain(self.output_concepts, [])
         comp_grain = Grain()
@@ -66,7 +69,7 @@ class GroupNode(StrategyNode):
                 len(parent_sources) == 1
                 and LooseConceptList(concepts=parent_sources[0].output_concepts)
                 == self.output_lcl
-            ):
+            ) and isinstance(parent_sources[0], QueryDatasource):
                 logger.info(
                     f"{self.logging_prefix}{LOGGER_PREFIX} No group by required, returning parent node"
                 )